All in the mind? Estimating the effect of mental health on health behaviours

Abstract

BackgroundHealth behaviours and mental illness have been found to be strongly correlated, both across and within individuals. However, establishing the precise nature of this link, particularly in terms of causality and its direction, remains a challenge. Much of the relevant existing research follow an experimental design and focus on measuring how changes in health behaviours affect mental health. ObjectiveThe study examines changes in mental health and health behaviours within individuals across time and evaluates the evidence for a causal pathway from psychological distress to risky health behaviours. MethodUsing a population representative longitudinal survey providing data from Australian households, this study employs a fixed-effects panel model framework to account for unobserved time-invariant individual heterogeneity. An instrumental variable estimation strategy is employed to specifically evaluate the evidence of a causal pathway from psychological distress to risky health behaviours. ResultsThe results confirm strong within-individual associations between psychological distress and health behaviours. Further evidence supports a causal relationship from psychological distress to health behaviours, though only for men and lifestyle habits. The case for a direct causal pathway is less straightforward when considering dietary habits, and potentially also for women. This lack of clarity suggests at least some of the observed within-individual associations reflect causal effects of psychological distress on risky health behaviours. ConclusionsGiven the substantial burden of disease associated with unhealthy behaviours, further research examining potential causal pathways from mental health to health behaviours should be a priority, as there is the potential to reduce health burden through mental health improvements.